Hidden Voltage Regulator

In the past 43 years I've built 4 solid state regulators, and one of them was on AC but still had to run DC to the field. The 4th one I built twice because AC or reversed polarity (its on a Chinese steam locomotive at the Boone and Scenic Valley RR) burned up the parts. It wasn't so compact that it would hide inside a switch box. And I've designed and built an AC current regulator using a DC IC inside a bridge rectifier. Making the design for AC adds considerable complexity in needing a bridge rectifier for the sensing and power supply, and then for the field control element with significant need for isolation between the sensing and the control segments of the circuit. Then there is the added complication of two ways to wire a generator or alternator field, though nearly all alternator fields are grounded at one end and fed power while generators are split between being grounded and being connected to the armature terminal which reverses the current in the control element. One way the control element (transistor probably) has to supply power, the other way the control element has to supply a variable resistance ground.

Its conceivable that a circuit could be built that could tolerate reversed polarity (but would not work with reversed polarity) by adding some series diodes and that could be wired with jumpers for either polarity. Handling both field connections essentially means a duplication of the control transistor because driving that transistor without a higher voltage than the generator output voltage means a different polarity transistor for high side than for low side control. The added series protective diodes will add power dissipation and temperature sensitivity to the voltage sensing circuit and may limit the maximum output or the lowest speed for charging of the generator/alternator. Simply because the voltage drop in the diode is temperature sensitive and reduces the voltage the control element can apply to the field coil. Then there's the complication of getting the inductive kick suppression diode correct for either generator field connection and polarity. And having the circuit survive the flashing necessary to get the generator to build with the same polarity as the battery.

A formidable design task. Not easily accomplished for free.

Gerald J.

Thanks for the input, was over simplifying this. I had not done any research yet but after messing with my voltage regulator thougt there would be a better way with electronics. The over simplify is that am only looking at the field circuit and forgot about building the initial field.  

The LM317 and LM340 family have the ruggedness required for the inductive kicks of the field coils and the voltage kicks associated with alternators. Unfortunately they aren't arranged in a way that works for regulation the field. They take an unregulated voltage and pass it through with a constant output voltage, so long as the input is more than a couple volts over the output voltage and under about 35 volts and they don't get too hot. They are nice for that application with current limiting and thermal limiting. I'm sure everything solid state item I have uses several.

I've not looked for a solid state alternator regulator, though there should be some because I've bought solid state modern regulators in metal packages to emulate the magnetic vibrating regulators. Early solid state factory regulators used a zener diode for reference and two or three bipolar transistors.

Yes a chopping regulator in solid state would be nice for regulator power dissipation, probably nastier on radio reception in the car or truck but thyristors are unsuitable for DC because they aren't easy to turn off. GTO devices or power MOSFETS are nicer. I like power MOSFETs. I used a power MOSFET in my steam locomotive regulator with a TL31C set up to shunt the gate drive to regulate. I use the TL431C open loop with a feedback capacitor to slow it down to the speed of the MOSFET. There is a TL431C in virtually every desktop computer supply these days. So the part is not hard to buy. Select a pair of resistors and it makes a controller for any voltage from 2.5 to 35 volts, though I know of circuits regulation 350 volts for radio transmitters based on that little chip. When they burned up my first steam locomotive regulator from AC or reversed polarity, the TL431C survived.

Getting the generator to build when started is a concern. In mine I used a fixed power resistor across the control transistor to set a minimum field current. Then with any voltage rise, I turned on the power MOSFET and turned it off with the TL431C. That power resistor in the RR regulator (32 volts) takes up more volume than the rest of the regulator.

I've considered, but not designed yet, that many a PIC chip has an analog input or as many as there are I/O pins and has the capabililty of a pulse width modulated output. That would probably be the device to make a pulse switching generator voltage regulator. It needs 3 to 5 volts supply, and will idle on a few microamps, so a 2032 lithium cell will run it for years. Some cost as little as 79 cent each, it would take a bit of software to be embedded to look at the voltage samples and to change the pulse width drive to the external power device (probably a MOSFET) to control the detected voltage. It would need a reset on engine start which might be detected from ignition switch power. I think there are some fancy switching supplies already doing that. It would only rest on the lithium cell, not run on it. Many can supply their own processor clock. And there are competing devices from many other makers. Its a major task to learn each maker's programming schemes and I like PIC chips so much I've bought a chunk of stock in the company. So I'm biased. There are hundreds of PIC chips to choose from.

Now is there enough market to cover weeks of engineering and testing?

Gerald J.

Another source of small regulators are the ones for B&S engines,most are good for 15 amps,based on thyristors( I've 'depotted' them).Only good for negative ground systems though as the case must be ground.

Here is a site for the Podtronics regulator used for the Lucas generator on cycles , it is set for the ground you wish to use . It can be placed inside the case of a present regulator to hide the New unit for those who want original look 

Edited by Coke-in-MN - 23 Sep 2012 at 11:13am

Brett, It would be nice if they could cram that all in a cutout relay box, but there aren't enough lugs sticking out It would be lot easier than trying to figure out what your battery needs while you are playing in the dirt